Plant-induced defenses triggered by herbivory generally decrease insect herbivore performance.  Such defenses, however, can also interfere with the transmission of fatal insect baculoviruses, leading to an increase in herbivore fitness.  Because of these effects, induced plant defenses can change both the short-term and long-term dynamics of insect outbreaks.   Previous work, however, has used laboratory dose-response experiments, which usually measure average transmission rates.  The theory of insect population dynamics in contrast has emphasized the importance of heterogeneity in transmission, which can only be accurately measured in the field.  To examine the potential effects of plant-induced defenses on outbreaks, we used jasmonic acid (JA) to increase tannin production in red oak Quercus rubra, a preferred food source for the larvae of the outbreaking gypsy moth Lymantria dispar.  We then used field experiments to see how tannin induction affected the transmission of the gypsy moth baculovirus. 
Results/Conclusions

Our results showed that tannin induction has only a modest effect on average transmission rates, but heterogeneity in transmission on experimental trees was strongly reduced relative to control trees.   Inserting our experimental estimates of transmission heterogeneity into insect host-pathogen models showed that heterogeneity in the controls was high enough to produce stable population dynamics.  Extending the models to allow for tannin induction, however, produces outbreaks that match outbreak dynamics in the field.  While much debate in ecology focuses on whether natural enemies or resource quality control animal population dynamics, our results suggest insect outbreaks are driven by interactions between natural enemies and resource quality.